I used the following command to gather summary metrics for my bam file generated via bowtie2 (tophat to be specific):
java -jar /usr/share/picard-tools-1.136/picard.jar CollectAlignmentSummaryMetrics INPUT=Sample_DY10.tophat.bam OUTPUT=tmpmetrics/alignmentmetrics R=/mnt/storage/ref_genome/Homo_sapiens/UCSC_hg19/UCSC/hg19/Sequence/Bowtie2Index/genome.fa
The output file is here: https://drive.google.com/file/d/0B2tk2ztVP7NIa1NNM2FrMi1WSjg/view?usp=sharing
The question I have is that the metrics PF_HQ_MEDIAN_MISMATCHES has a very high number (66). When I look at NM tag in the bam file, I see that the median is 1 with max NM = 2 I am wondering how this number is calculated by PICARD.
Any help is appreciated.